1. Kay-Uwe (Kiwi) Lorenz
  2. aptk

Overview

aptk - A Parse Toolkit

This is a module for creating parsers from grammars. aPTK targets to make this in a very "documented" way. So you can define the grammar in doc-string of your grammar class or even in your reStructuredText documentation.

A simple example of a greeting parser:

:grammar GreetingGrammar

<greeting>        := <greeting-clause> <.ws> <greeted>
<ws>              := \s+
<greeting-clause> := "hello" | "hi"
<greeted>         := \w+

This will create a class named GreetingGrammar. First rule in the grammar will be used as start-rule for a normal parsing. Parsing a rule, will result in a ParseTree. Here follows an example of rule <greeting> matching "hello world":

<greeting> ~~ "hello world"
           -> greeting(
                greeting-clause( 'hello' ),
                greeted( 'world' )
              )

Actually what you see above is a test assertion for applying rule greeting to string "hello world", what is expected to result in the parse-tree displayed above.

But read more in aptk's documentation.

Download

You can download this package from PyPI_.

Or install it with easy_install:

easy_install -U aptk

or get the the source from source code repository from bitbucket.org:

$ hg clone https://bitbucket.org/klorenz/aptk

Building Documentation

For building documentation you need sphinx, you can get it using:

$ easy_install -U Sphinx

Then you can:

$ cd docs
$ make html
$ firefox _build/html/index.html

License

Licensed under New BSD License, see LICENSE.txt.

Release Notes

Version Notes
0.5.7 add comments
0.5.4 first tests for postcircumfix pass, more tests, cleaned code a bit
0.5.3 Added operation precedence parser. Pretty simple yet, but supports infix, postfix, prefix, circumfix operations. postcircumfix is also implemented but not tested at all others are partly tested.
0.5

Starting with version 0.5, which shall indicate, that there is still a lot to do.

Especially documentation is far from complete and it contains some passages, which are already outdated.

Next releases will focus on documentation and testing and adding an operator precendence parser.